observed otus by DADA2, taxonomy

[anirban.mcgill (Anirban Kundu)]

Hello,
I have 2 questions (1 regarding DADA2 and 1 regarding taxonomy results). They are short questions so I put them in the same post. Thanks for your help!

I ran the DADA2 denoise pipeline followed by the core-metrics-phylogeny command to obtain alpha and beta diversity indices. I noticed that the observed-otu command gives number of OTUs (in this case ASVs because I am using DADA2) between 80 to 120 for my samples. My samples are soil samples, I was wondering if it was too low a number?

I ran my taxonomic analysis using the 99_otus fasta and taxonomy file from the gg_13_8 database. I retrained my classifier with my primer set (also because the scikit package was incompatible). I noticed there are some taxonomic annotations with no information from the phyla level down to the species level. (E.g. k_Bacteria;p__;c__;o__;f__;g__;s__)
The number of these annotations are really low, yet I was wondering why so. Are these the unknown samples?

Best,
Anirban

[timanix (Timur Yergaliyev)]

Hi!

@anirban.mcgill

I noticed that the observed-otu command gives number of OTUs (in this case ASVs because I am using DADA2) between 80 to 120 for my samples. My samples are soil samples, I was wondering if it was too low a number?

What kind of soil?
I analysed rhizosphere soil samples and got number of observed ASVs in range from 750 to 1250. But if your samples are soil form the dessert, where no plants are growing, the number may be much lower. I think you should check relevant papers.

@anirban.mcgill

noticed there are some taxonomic annotations with no information from the phyla level down to the species level. (E.g. k_Bacteria;p__;c__;o__;f__;g__;s__ )
The number of these annotations are really low, yet I was wondering why so.

I always getting them as well. I believe there are several reasons why the classifier is failed to annotate them deeper, including your guess. You can just filter them out.

[Nicholas_Bokulich]

@anirban.mcgill

I noticed there are some taxonomic annotations with no information from the phyla level down to the species level. (E.g. k_Bacteria;p__;c__;o__;f__;g__;s__ )
The number of these annotations are really low, yet I was wondering why so. Are these the unknown samples?

@timanix

I believe there are several reasons why the classifier is failed to annotate them deeper

These are not due to the classifier failing to annotate, they are the annotations given by your reference database. See this topic for details:

ASV collaps question

Hi Collaps to genus level in qiime2. It is divided into one with and without a level name. Can't you see the same bacteria? Thank you.[%EC%BA%A1%EC%B2%98]